Why we need a philosophy of engineering:

a work in progress
STEVEN L. GOLDMAN
Departments of Philosophy and History, Lehigh University, Bethlehem, PA 18015, USA

Engineering problem solving employs a contingency based form of reasoning that stands
in sharp contrast to the necessity based model of rationality that has dominated Western
philosophy since Plato and that underlies modern science. The concept ‘necessity’ is
cognate with the concepts ‘certainty’, ‘universality’, ‘abstractness’ and ‘theory’. Engineering
Published by Maney Publishing (c) IOM Communications Ltd

by contrast is characterised by wilfulness, particularity, probability, concreteness and prac-

tice. The identification of rationality with necessity has impoverished our ability to apply
reason effectively to action. This article locates the contingency based reasoning of engineer-
ing in a philosophical tradition extending from pre-Socratic philosophers to American
pragmatism, and suggests how a contingency based philosophy of engineering might enable
more effective technological action.

For reasons that have been at the heart of Western culture from its beginnings in
ancient Greece, engineering has been treated dismissively by intellectuals and in particular
by philosophers.1 The reasons reflect deeply rooted prejudices that have been sustained
for well over two thousand years. Of special relevance to the persistent underestimation
of engineering is the low value historically placed by intellectuals on the contingent, the
probable, the particular, the contextual and the temporal. Conversely, philosophers
especially have placed a high value on the necessary, the certain, the universal, the context
independent and the timeless. This hierarchisation subordinates practice, values, emotion
and will to theory, value neutral principles and deductive logic in ways that leave us ill
equipped to deal rationally with life. Indeed, from the perspective of necessity, life and
action are fundamentally irrational!
Engineering is paradigmatic of what is undervalued in Western ‘high’ culture, that is, in
the culture of ideas, of education, of art, of morality (and of the monotheistic religions).
Engineering is contingent, constrained by dictated value judgements and highly particular.
Its problem solutions are context sensitive, pluralistic, subject to uncertainty, subject to
change over time and action directed.
By contrast, mathematics is paradigmatic of what has been most admired in Western
‘high’ culture, namely reasoning that is abstract, necessary and value free; and problem
solutions that are universal, certain, unique and timeless. Historically, ‘demonstration’,
meaning mathematico-deductive argument, is the form of reasoning that the most
respected Western philosophers from Plato and Aristotle to the early Wittgenstein have
striven for, rejecting reasoning based on the probable, the concrete and the contingent.
Reasoning ‘in the geometric manner’ is also the mantle in which modern science was
cloaked from Descartes and Galileo through Einstein and Schrödinger. By the twentieth
century, science had won growing public acceptance as the paradigmatic application
of reason to experience, and thus as uniquely capable of disclosing the truth about reality.
This acceptance is indebted at least in part to science’s use of esoteric mathematics and

DOI 10.1179/030801804225012572 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 163
© 2004 IoM Communications Ltd. Published by Maney for the Institute of Materials, Minerals and Mining
 a hybrid experimental logic that seems deductive (but is in fact an instance of the ‘fallacy
of affirming the antecedent’, as was noted in the seventeenth century).
The price we have paid for this identification of knowledge and truth with necessity and
its cognate concepts is a radical divorce of reason from action, one manifestation of which
is valuing science more highly than engineering. This divorce was already considered a fact
for ancient philosophers, and it remained influential throughout the twentieth century.
Aristotle argued that because action entails contingency, particularity and uncertainty, there
can be no ‘science’ of action.2 Knowledge/science cannot determine action because what
is meant by ‘knowledge’/‘science’ is the necessary, the universal and the certain. This
creates a gulf between theory and practice that cannot be bridged by deductive reasoning,
which alone is necessary and alone is able to achieve certainty. By default, then, action is
ultimately wilful, and determined by desire, which is essentially irrational.
In the course of the nineteenth century, various cultural critics attacked this conception
of rationality, and in the twentieth century a number of important philosophers explored
contingency based interpretations of rationality. But with few exceptions, among them
Published by Maney Publishing (c) IOM Communications Ltd

John Dewey, the spectacle of contemporary technological accomplishments did not

suggest to these thinkers that in engineering, an effective contingency based model of
rational action was being manifested. On the contrary, the development of bold new
scientific theories, as well as innovations in mathematics and logic, reinforced the prejudice
that necessity based rationality as exemplified by science was the key to understanding
experience and disclosing reality. Engineering continued to find its place as science’s
handmaiden.

ENGINEERING V. SCIENCE
Engineering is commonly described by the scientific community as applied science, and it
is accepted as such not only by the general public, but even by the engineering community
and its leaders. This characterisation of engineering was built into Science: The Endless
Frontier, the 1945 report to the President that transformed post-war US science and
technology policy to global effect.3 The report was written by Vannevar Bush, who had
been an electrical engineering professor at MIT and later President of the Carnegie
Institution before being made head of the Office of Scientific Research and Development
(OSRD) by President Roosevelt in 1940. It was the extraordinary wartime success of
OSRD that made Bush’s case for reversing the historic US policy of not using public
funds to support ‘pure’ science.
Bush was well aware that many of what were promoted as OSRD scientific triumphs,
including the atomic bomb, radar and electronic countermeasure devices, mass production
of penicillin and blood plasma, and the pioneer electronic computer ENIAC, were at least
equally engineering triumphs. But he was also astute enough politically to appreciate the
greater cultural prestige attached to science. Furthermore, as engineering was by 1945
overwhelmingly incorporated into profit driven enterprises, using public funds to support
engineering related activities conflicted with the ethos of industrial capitalism. Science: The
Endless Frontier thus presented a scenario in which public investment in ‘pure’ scientific
research was necessary because this disinterested research, aimed only at understanding
nature, alone generated the kind of knowledge that made available to engineering and
industry could create engines of economic growth and anchor future national security.
Bush also was well aware that the ‘real world’ science–engineering–innovation process
was far more complicated than this linear model represented it as being. Historically,

164 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 engineering and technology had led scientific theory at least until the second half of
the nineteenth century. At that time, innovations in the chemical, electrical, transportation
and communication industries were systematically coupling science and engineering,
motivating the creation of proprietary industrial research laboratories on the one hand and,
in Germany, publicly funded research institutes on the other. But in almost every instance
of innovation, from the steam engine and the telegraph to the photocopier and the
computer, creative non-scientists triggered a form of positive feedback between research
and innovation: commercially successful innovations stimulated new science, which
enabled new engineering, which led to improved or new applications, which drove further
research and newer innovations.
Bush’s report was directly responsible for the creation, after bitter political controversy,
of the US National Science Foundation (NSF), which rejected the very idea of engineering
research right into the 1980s. The NSF position was that research meant new knowledge,
and knowledge was solely the purview of science. Engineers merely applied knowledge in
practical ways, which implied that any intellectually interesting issues posed by technical
Published by Maney Publishing (c) IOM Communications Ltd

knowledge were in the domain of science, not engineering. Philosophy of science, for
example, quickly became a respected subdiscipline of philosophy, pursued since the
nineteenth century by leading philosophical and scientific thinkers. It has been comple-
mented in the twentieth century by the rise of history of science and sociology of science
as scholarly disciplines. Philosophy of engineering, by contrast, is virtually unknown in
the Anglo-American world, and history and sociology of engineering are marginal
subspecialities, at best.4
Starting in the early 1960s, academic science, technology and society (STS) programmes
have had a significant effect in making large numbers of students aware of environmental,
political and ethical issues associated with the social impact of technology.5 But except for
courses in engineering ethics, engineering is treated in almost all the courses that make up
such programmes as a black box, as a step in the technological innovation process, or, in
case studies, as technical problem solving. Engineering as it is practised in commercial or
governmental contexts is rarely addressed. STS scholarship, however, has generated a
vastly richer understanding of technological innovation as a complex social process. The
process is one in which technical knowledge is selectively exploited on behalf of institution
specific agendas driven by commercial and/or political values.6 It is the way that engineers
function as enablers of this process of selective exploitation of technical knowledge that
needs to be understood in order to appreciate engineering as exemplifying a distinctive
form of rationality vis à vis science.
Whatever the reasons for the low cultural esteem in which engineering is held, the
consequences for society are profound.7 Technological innovation continues to be a
primary agent of social change, as it has been since the industrial revolution began in the
late eighteenth century. The power of technologies, which increased at an accelerating rate
throughout the twentieth century, poses increasingly serious social, political and environ-
mental challenges. Our responses to these challenges have been woefully inadequate,
reflecting a preoccupation with arguing the respective merits of competing moral, ethical,
political and philosophical universal principles. Because after two thousand four hundred
years, mainstream Western philosophy has still not reached a consensus on what these
universal and necessary social, ethical and political values are, technology policy debates
dissolve into ideological conflict.
In fact, the situation is far worse than this. Technological action is only one instance of
profoundly threatening global challenges posed by economic, social, political, cultural and

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 165

 religious action. All forms of action directed policy debate, not just technology policy,
suffer from the gulf between theory and practice, knowledge and action, that even
Aristotle had identified as a consequence of embracing a necessity based conception of
rationality. Understanding engineering reasoning will not resolve these problems, but it can
lead us out into an examination of philosophies of the contingent and the concrete that
could point the way to solutions.8

HOW IS ENGINEERING DISTINCTIVE?

The definition of engineering problems, as well as of what will count as acceptable
solutions to them, explicitly depends on highly contingent value judgements that are
external to the technical expertise engineers command. These value judgements derive
from the projected economic, social and/or political consequences of the implementation
of solutions to engineering problems. The assessment of these consequences in turn
reflects the fact that engineering practice always takes place within highly specific, com-
Published by Maney Publishing (c) IOM Communications Ltd

mercial and/or political action contexts. Thus engineers, in order to function as engineers,
must have a boss, or at least a client.9 Scientists, on the other hand, are perceived as disin-
terested pursuers of universally true knowledge of the way things are. Their problems are
given by Nature, not by their employers, and Nature is the sole arbiter of correct solutions.
Typically, engineers solve problems for enterprises whose management is already
committed to specific courses of action and who employ engineers to enable those courses
of action. Neither the definition nor the solution of science problems, by contrast, is
dependent on action – though research may be motivated and funded by an action agenda,
as in the case of nuclear science in the Manhattan Project – and value judgements external
to the methodology of science are prohibited from a role in defining problems or
proposing solutions.
Obviously, engineers use mathematical and scientific knowledge to solve their problems,
but they do so in ways utterly different from the ways that mathematicians and scientists
solve their problems. Engineers use mathematical and scientific knowledge in ways that
are analogous to scientists’ use of mathematics and technology when solving scientific
problems, namely, on their own terms. For physicists, mathematics is a source of concep-
tual ‘tools’ to be used opportunistically to solve physics problems to the satisfaction
of physicists. Mathematicians may be dismayed by the way ‘their’ mathematics is so used
and might not accept as solutions to problems in mathematics the solutions accepted by
physicists to physics problems, but that is irrelevant to physicists.
A similar situation exists with regard to engineers’ use of scientific theories and mathe-
matical techniques. These serve as conceptual tools and techniques to be used opportuni-
stically by engineers on engineering’s terms. In addition, when engineers use materials
from science and mathematics, the universalistic character of these materials must be
adapted to the particularity of engineering problems. Engineering is thus no more applied
science than physics, for example, is applied mathematics.
There is a profound difference between engineering design and scientific theorising that
further undermines the characterisation of engineering as applied science. While at any
point in time there may be rival scientific theories of some phenomenon, in principle there
can be only one theory that is ‘true’, namely the uniquely correct account of the way things
are ‘out there’. Design, however, is an irreducibly pluralistic exercise of reason because of
the role played in design by contingent value judgements, which from the perspective of
working engineers often appear arbitrary. These contingent value judgements – embodied

166 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 in performance specifications and specification of size, weight, production cost, reliability,
materials, time to market, manufacturability, serviceability – determine the parameters in
terms of which both engineering problems and what will be recognised by management as
acceptable solutions to them are defined. Furthermore, designs are open ended: they evolve
over time as problem and solution parameter weights vary.
Design is thus a contextual and a historical process as well as being intensely particular.
Scientific theories, on the other hand, if correct do not evolve; ideally they are closed and
unique. Where scientists aim at the truth about nature, engineering design reflects what
Herbert Simon, describing managerial decisionmaking, called ‘bounded rationality’ and
‘satisficing’ – consciously operating under conditions of partial information and acting on
solutions judged good enough to do the job that needs to be done, even though they are
not optimal.10
While engineering problems are explicitly action directed and driven by value judge-
ments, scientific theories are explicitly value neutral and their purpose understanding. The
whole point of the seventeenth century methodological ‘revolution’ in the study of nature
Published by Maney Publishing (c) IOM Communications Ltd

was elimination of the person of the subject of knowledge from the knowledge itself. This
had as a corollary effect eliminating any necessary connection between knowledge so
gained and action by the subject. Scientific knowledge is equivocal with regard to action.
That is, what we are to do with scientific knowledge when achieved cannot be a question
for scientific knowledge. Technology, however, is intrinsically action directed. Auto-
mobiles are for driving, while a theory of the nucleus is a theory of the nucleus, not ‘for’
building a bomb or a nuclear reactor or anything else for that matter, except incorporation
into a wider theory.
Basing action on scientific knowledge requires supervenient value judgements, adding
judgements to science that come from outside science, for example from governments
and entrepreneurs. At the birth of modern science, Francis Bacon and Descartes had
proclaimed that scientific knowledge would give us power over nature with which to
improve the human condition, but long before the seventeenth century was over it
had been recognised that scientific knowledge and improving the human condition were
disjoint enterprises!11

CONTINGENCY IN WESTERN PHILOSOPHY

Reflecting on the distinctiveness of engineering vis à vis science suggests a wider distinc-
tion within Western ‘high’ culture between two clusters of cognate concepts, organised
here (see table overleaf ) under two covering principles: the familiar principle of sufficient
reason (PSR), and what I have called a principle of insufficient reason (PIR).12 These
concept clusters are neither unique, nor exclusive, nor exhaustive. They are offered here as
being suggestive of two modes of reasoning, of two different conceptions of what it
means to give reasons and to be reasonable, and of what will constitute knowledge and
truth. The cluster of concepts listed under PSR, in which necessity seems to me to play a
pivotal role, is the preferred one in mainstream Western philosophy. It is critical to what
‘rationality’ means in the dominant philosophical-mathematical-scientific tradition. The
concepts listed under PIR, on the other hand, in which contingency plays a comparably
pivotal role, are commonly associated by PSR with sophistry, rhetoric, scepticism,
historicism, psychologism and relativism, all having pejorative connotations.
The Western philosophical tradition has been dominated by a necessity based model of
rationality expressive of PSR that is epitomised by mathematics and exemplified in modern

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 167

 Two clusters of cognate concepts: the principles of
sufficient reason (PSR) and of insufficient reason (PIR)
PSR PIR

Intellect Will
Reality Experience
Knowledge Belief
Truth Opinion
Certainty Probability
Objectivity Subjectivity
Universality Particularity
Absolute Relative
Necessary Contingent
Deduction Induction
Abstract Concrete
Theory Practice
Contemplation Action
Published by Maney Publishing (c) IOM Communications Ltd

Understanding Use
Prediction Anticipation
Unique Plural
Closed Open ended
Timeless Historical
Utopian Contextual

science. What is subjugated in this tradition is a contingency based model of rationality

that is expressive of PIR and exemplified in engineering. For a philosopher working with
PIR concepts, knowledge, truth and certainty as defined under PSR simply do not exist.
They must be redefined consistent with belief, opinion and probability if they are to refer
to anything actual. For philosophers working under PSR, the concepts associated with PIR
are what must be transcended in order to achieve knowledge, truth and certainty.

ENGINEERING AND SCIENCE: TWO RATIONALITIES?

The rationality of engineering can be described as being different from the rationality of
science in exactly the same way that the rationality of contingency based philosophy is
distinct from the rationality of necessity based philosophy. It is particularly significant for
identifying these as two different conceptions of rationality that the fact/value distinction
is central to so called ‘hard’ scientific reasoning but impossible for engineering reasoning;
and that engineering problem solving intrinsically anticipates action, whereas scientific
problem solving does not.
The contrast between the PSR and PIR concept clusters does as a matter of historical
fact reflect an open ‘war’ between two competing conceptions of rationality in the history
of Western philosophy. This war begins with Plato’s attack in his Gorgias on the Sophists13
as abusers of reason, teachers of tricks for winning arguments who do not know what
the good, the right and the true are. They are thus not philosophers at all, not lovers
of wisdom as Socrates was. But the Sophists, in the words of historian Nancy Struever,
deliberately chose ‘to shun the ideal sphere where pure reason and perfect justice reside’,
which for Plato was the objective of philosophy, ‘for the shifting and uncertain field of
action and discourse’.14 The Sophists denied the reality or even the possibility of absolute

168 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 truth and absolute value, denying as well the superiority of the universal to the particular
and of abstract theory to concrete practice.
For the Sophists, the goal of philosophy was in the first instance illuminating action.
For them, the way to philosophy was through rhetoric, which was not only techniques
of persuasive speech, but the discovery through social speech of how to act well. Plato’s
desire for ‘purity of thought and communication’ was for the Sophists a delusion, and even
the power of deductive logic to compel assent was for them (Struever again) ‘mediated
through the passions, not just the intellect’. Protagoras’ claim, mocked by Plato,15 that
‘Man is the measure of all things’ thus reflected a view of philosophy as rooted in experi-
ence, not in an unexperienced and unexperiencable ‘reality’ that transcended experience.
As such, the goal of discovering the best way for a person to act must proceed from an
understanding of how people act, of how Man ‘measures’ things, how people, individually
and in groups, assign values to their experiences.
The triumph of the Platonic–Aristotelian interpretation of philosophy not only subordi-
Published by Maney Publishing (c) IOM Communications Ltd

nated action to understanding, but was couched in terms that trivialised a concern with
action as ignoble. Isocrates, a contemporary of Socrates and a student of Gorgias, argued
passionately but unsuccessfully against this interpretation of philosophy on the grounds
that experience is in fact contingent, particular and uncertain. Decades before Aristotle,
Isocrates saw that basing philosophy on necessity, universality and certainty entailed an
abstraction from experience that made such knowledge useless for action. In the Antidosis,
Isocrates turned the already pejorative term ‘sophist’ against Plato. It was Plato, he said,
who was the sophist, teaching intellectual gamesmanship that gave us no help in making
decisions, not the rhetoricians. The rhetoricians were the true philosophers, pursuing
practical wisdom.
In his Rhetoric and Nicomachaean Ethics,16 as well as in On the Soul, Posterior Analytics and
Politics, Aristotle acknowledged a gulf between theoretical wisdom and practical wisdom
that cannot be bridged by reason. Rhetoric, he writes, is an offshoot of dialectics applied
to political science and ethical studies. ‘There are few facts of the necessary type that
can form the basis of rhetorical syllogisms. Most of the things about which we can make
decisions and into which we inquire present us with alternative possibilities. For it is about
our actions that we deliberate and inquire, and all our actions have a contingent character;
hardly any of them are determined by necessity.’ All action based reasoning is rhetorical
and as rhetoric and dialectic are practical faculties and essentially contingent, they cannot
be sciences and therefore there cannot be a science of action.17
Rhetoric and dialectic, Aristotle noted, are the only arts of reasoning that draw opposing
conclusions, in principle in order to explore vulnerabilities of probable arguments in
pursuit of the ‘right’ conclusion.18 Of course, this opens the possibility of abuse of rhetori-
cal and dialectical skills if techniques of effective persuasion are separated from guiding
values. Over two thousand three hundred years ago, then, what would become the core
issue of twentieth century critics of technology had already been raised: that means have
overwhelmed ends, that a fascination with technique has overwhelmed any understanding
of why to employ technique, how to employ it and to what ends. On the Aristotelian view,
engineering is allied to rhetoric, which lacks understanding and rationality but promotes
action. Science possesses understanding and rationality, but is equivocal with respect to
action!
That necessity and contingency were ‘at war’ intellectually as bases of competing
conceptions of rationality is fundamental to Classical scepticism, which rests on a rejection

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 169

 of a necessity based conception of knowledge and of reason. In his Academica, written in
the first century BCE, and an important influence on Renaissance thought, Cicero defended
the ‘cleansing’ of philosophy of the concepts ‘necessity’ and ‘certainty’. The sceptical phi-
losophers wanted knowledge and reason defined in terms of contingency and probability
because these are facts of experience. No one, they noted, has yet provided a criterion for
identifying universal, necessary and certain knowledge and truth. Philosophical scepticism
is thus a call for a recovery of what we mean by the terms ‘knowledge’ and ‘reason’, from
the self-proclaimed ‘true philosophers’ who have dogmatically co-opted the right to define
those terms.
Sceptical philosophy was unacceptable to Christianity – Augustine, for example, once he
became a Christian considered it imperative to refute scepticism – because Christianity
preaches an absolute truth. By the late medieval period, however, the primacy of necessity
and universality in philosophical and theological reasoning was being challenged within
Christian universities. In the fourteenth century, Duns Scotus and William of Ockham
Published by Maney Publishing (c) IOM Communications Ltd

made will and particularity central to rationality. The political no less than the intellectual
and theological controversies precipitated by this shift from universal to particular and
from logic to will were intensified by the Renaissance revival of Cicero’s writings in the
fifteenth century and in the sixteenth century by the translation into Latin of Classical
sceptical philosophical manuscripts.19 The nature of rationality and its limits, the respective
claims of certitude and probabilism, were critical issues for the Protestant Reformation
and the Catholic Counter-Reformation as well as for Renaissance philosophy.
It was in the Renaissance that contingency made its strongest assault on necessity as the
basis of philosophy and rationality. The revival by the Humanists of rhetoric and of history
as the ‘true’ bases of philosophy precisely because of their focus on action and embrace
of particularity and contingency was an open declaration of cultural ‘war’. ‘Rhetorical
concepts of discourse emphasize change . . . the many . . . the particular . . . emphases which
are essential in a serious commitment to historical understanding, i.e., historicism.’20 And
indeed the Humanists invented historicism, which carries with it contextualism, relativism,
pluralism and openendedness.21 Rhetoric, like engineering, is coupled to action, to deci-
sionmaking in order to act, and to the making of distinctions in order to ‘rationalise’
decisionmaking. Also like engineering, rhetoric’s goal is ‘manmade stability in an unstable
world of relationships’. Not surprisingly, the roots of modern engineering also lie in
the Renaissance. The Humanists recovered, edited and published numerous Classical
engineering texts, among them Vitruvius’ On Architecture, books on mechanics by Philo
of Byzantium and Hero of Alexander, and applied mathematical works by Archimedes.
Concurrently, the design of more complex and more capable machinery was enabled by
the invention of engineering drawing – cutaway machine drawings, ‘exploded’ views of
parts, orthogonal projection – itself indebted to the spread among artists after 1450 of
central vanishing point perspective drawing techniques. This was coupled to intensified
industrial and commercial activity, leading to increasingly complex military and civilian
technological projects.22
The Humanist assault on necessity was rebuffed, of course. It was defeated by the over-
whelming success of explicitly antisceptical, necessity based seventeenth century Rationalist
philosophies such as those of Descartes, Spinoza and Leibniz, by the rise of modern
science and its materialist-determinist interpretation of nature, and by the eighteenth
century apotheosis of Reason as the ultimate value and even the ultimate reality.
Hume’s attempt to formulate necessity free theories of knowledge and ethics were crushed

170 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 by the necessitarian philosophies of Kant and Hegel and by the growing power of the
cult of science. In the nineteenth century, however, the Romantic literary revolt against ‘the
Age of Reason’ and ‘soulless’ science was complemented by critiques of necessity based
philosophy in work by Schopenhauer, Kierkegaard, Nietzsche and Bergson that emphasised
the primacy of will and action, not abstract reason.
From its origins in the 1870s, American pragmatism was an action centred philosophy
that made contingency central to its conception of rationality and to its conceptualisation
of experience. In his Quest for Certainty, John Dewey explicitly contrasted necessity and
contingency as rival conceptions of rationality and competing bases of philosophy. Unlike
the Romantics and the philosophical critics of reason, however, pragmatists generally and
Dewey especially were enthusiasts for science, the scientific method and the application of
scientific reasoning to all aspects of life. But for Dewey, the key to understanding science
lay in engineering!23 Dewey argued that science was a form of engineering, that it was only
hypothetically abstract, universal, necessary and certain. In truth science was as value laden
Published by Maney Publishing (c) IOM Communications Ltd

and ‘interested’, hence as contextual, as engineering.

For Dewey, the hypothesised necessity based rationality of modern science was an
idealisation of contingency provoked by experience. The experience of insecurity, of
vulnerability, of the ‘fragility of the human condition’ provokes religion as one response, a
philosophy that promises universal, necessary and certain truth as another, closely related
response and also the response of effective, because reasoned, action. Science and engi-
neering are thus not truly opposed to one another. There is one process in which
both inhere: the process of systematic correlation of action with its actual and intended
consequences. For cultural reasons, science has been misperceived as a form of necessity
based philosophy, which Dewey stigmatised as religion under another name.24

TOWARDS A PHILOSOPHY OF ENGINEERING

What is the point of this sketch of an outline of a history of the idea of contingency in
Western philosophy? What claims is this sketch intended to support?
First, that there have been rival conceptions of rationality from the beginning of Western
philosophy, between which open hostility has been the historic norm, proponents of each
routinely attacking proponents of the other as deluded, at best.
Second, that necessity and contingency have played the role of what the sociologist
Alfred Schutz called ‘key concepts’ in these rival models of rationality, that is, concepts
whose function, when clarified, simultaneously illuminates a cluster of correlated concepts.
Third, that the choice of basing rationality either on necessity or on contingency underlies
mutually exclusive philosophical theories of truth, knowledge, values/action and reality.
Fourth, that engineering has strong affinities with contingency based models of
rationality, that this strand of Western philosophy is the intellectual ‘home’ of engineering,
while necessity based models of rationality with which science has strong affinities have
dominated the Western tradition.
Fifth, that action poses a serious problem for the necessity based model because of a
disjunction between theory and practice, while action is incorporated into the core of the
contingency based model.
Finally, that philosophy is relevant to engineering; that understanding engineering fully
as a practice and as a form of reasoning requires appreciating engineering’s place within
a particular philosophical tradition; more, that it requires recognising that engineering is

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 171

 an instance of a particular model of rationality that is the nexus of a cluster of cognate
concepts that have implications for engineering reasoning and practice.
Engineering practice in the modern world is embedded in a very particular social
context, one that has evolved out of mid twentieth century industrial capitalism. This
context today is distributed globally in ways that ignore national boundaries, political
philosophies and social organisation. The common denominator is the process within
which engineers function. As alluded to above, this process is one in which engineering
serves managerial agendas. Engineers apply their expertise to the solution of problems
that derive from these (commercial or political or military) agendas, and their solutions
enable the realisation of these agendas. Engineering is thus ineluctably a sociopolitical,
as much as a technical knowledge, practice. As a matter of historical fact, engineers in
the Anglo-American world have overwhelmingly insisted that they are only technical
problem solvers, that accountability for actions based on their solutions and for conse-
quences of those actions lies with others. This insistence rings hollow, however, with
deeper insight into the nature of engineering and the cumulation of negative consequences
Published by Maney Publishing (c) IOM Communications Ltd

of technological action.
If engineering reasoning is by its very nature embedded in action contexts, then
engineers cannot escape sharing responsibility for that action. But it is only within the
framework of an action philosophy that we can apply value judgements rationally to
the action that engineering, and engineers, enable. ‘Rational’ technology policymaking
and technology assessment inevitably elude intellectuals who come to these processes with
the prejudices associated with the necessity based intellectual-philosophical tradition – for
example, believing that the process requires identifying universal principles and values
from which ‘right’ technological action and ‘good’ engineering can be deduced; or
who believe that as this effort is hopeless, the process should take as its goal a ‘functional’
balance among the interested parties, whose respective interests are weighted in a way that
is misleadingly, often cynically, called ‘pragmatic’.
‘Pragmatism’ was an American philosophical innovation, initiated by Charles Sanders
Peirce in the 1860s and 70s, adapted and promoted by William James at the turn of
the century, but most systematically developed by John Dewey beginning in the 1890s. It
is pragmatism that seems to me to hold out the greatest hope for the rational assessment
of technological action and of engineering as its enabler; the greatest hope, too, for a
meaningful – and implementable – engineering ethics. ‘Instrumentalism’ functions as a
key concept in Dewey’s version of pragmatism: clarifying this concept simultaneously
clarifies a cluster of correlated concepts. In the necessity based/principle of sufficient
reason philosophical tradition, instrumentalism is a pejorative term, implying a focus on
means rather than ends, on getting some job done rather than on pursuing an understand-
ing of what ultimately gives that job value and meaning, thereby legitimating doing that job
in the first place. (After two thousand four hundred years with no consensus yet on such
an understanding, philosophers attribute the highest value to the pursuit itself, a move to
which Plato had recourse in order to justify Socrates’ inability to identify the universal
values he believed in.)
What Dewey means by instrumentalism, however, differs fundamentally from its
meaning in mainstream philosophy. For Dewey it is the name of the complex process
through which we respond deliberately and effectively to experience, a name for the
content of consciousness. Consciousness, in turn, is the selectively interested, actively
engaged, constantly evolving interactive process, ultimately intersubjective, that produces
experience. Mind, as subject of experience, and world, as object of experience, are both

172 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 intellectual constructs that misleadingly attribute thinghood to aspects of this process.
What is real is the process, the constantly changing process, of consciousness. The self-
active aspect of consciousness is reflected in the selectivity of our attention and in the
projection of structure, including the structures of closure, anticipation and control, onto
this continually changing, fundamentally contingent experience of self and world. More
broadly, the ultimacy of process – within which we make contingent distinctions based on
interests and to which we selectively attribute thinghood – is central to Dewey’s philoso-
phy. It was the basis for his opposition to all dichotomous, either/or thinking, to all
abstraction from experience of ‘atomic’ realities that exist outside the process with fixed
properties of their own. Mind attributes thinghood all the time, of course, to features of
the content of experience, but what we mean by each of these ‘things’, by space, time,
matter, energy, atom, gene, earth, universe, continually evolves as experience, and our
interests, do.
The active and interested character of consciousness simultaneously creates the context
of inquiry, that is, thinking focused on some facet of experience that appears as problem-
Published by Maney Publishing (c) IOM Communications Ltd

atic and that provokes a response to resolving that problematicity. Explicating the logic
of inquiry is the heart of Dewey’s pragmatism and it is to that logic that instrumentalism
refers. Inquiry is cognitive consciousness; it appears as a deliberate reflection on wanting
to change a particular state of affairs, either to end a present undesirable one or bring
about a desired but absent one.
Note that inquiry is intrinsically value laden, teleological, emotive and wilful, but also
logical. What makes inquiry and cognition rational for Dewey is the clarity of the determi-
nation and assessment of the end desired, the appropriateness of the means specified to
achieve that end, and the attention paid to the evolving experience of implementing
the specified means, which may require modifying, even abandoning, the end no less
than modifying the means. Consistent with his pragmatist framework, what we mean by
‘knowledge’, ‘truth’ and ‘reality’ involves beliefs keyed to the perceived effectiveness of
actions, which again makes values essential to these definitions.

CONCLUSION: GETTING THERE FROM HERE

Dewey’s instrumentalist conception of rationality seems to me precisely the rationality
of engineering, as Dewey himself recognised. His pragmatic philosophy can be the frame-
work for a philosophy of engineering, and a philosophy of action generally, one that does
full justice to the pervasively contingent realities of engineering practice. At the same time,
it can be extended beyond engineering to encompass the social process by means of which
engineering knowledge, and engineers, are selectively exploited in support of commercial
and political agendas that they have no hand in setting. Given the increasingly problematic
character of technological action for societies literally worldwide – with more and more at
stake physically, economically and culturally, and higher stakes, for more people, in more
countries – more effective responses to technology related problems are demanded than
the all too visible hand of corporate and entrepreneurial greed, national political agendas,
religious ideology or fatalism. As we move into an era of intensified innovation and activity
in a wide range of biotechnologies and nanotechnologies, even as existing environmental,
energy, food, water, health, wealth distribution and demographic problems worsen, a
philosophical framework within which to assess values critically is long overdue.
There is no sign that traditional philosophy can provide such a framework, but studying
engineering as a functioning real world implementation of Dewey’s pragmatism seems a
fertile approach to a starting point for developing such a framework. Dewey, however, left

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 173

 a glaring hole in his theory of experience: how to evaluate ends? If experience is all there
is, with no ‘beyond experience’ to validate what we mean by knowledge and truth, values
and meanings, and if ends are simply ‘there’ in experience, how can we judge whether we
should want what we find that we do want? How can we use the terms ‘good’ and ‘bad’
with reference to ends without falling into instrumentalism in the narrow, traditionally
pejorative sense in which the rationality of means can be studied critically but not ends?
How can experience be a closed and fundamentally contingent system of relationships
and still be a source of all of the values and meanings required by a consistent and
comprehensive philosophy of action. One approach to a solution to this problem may lie
in a twentieth century intellectual development that cuts across the scientific disciplines:
relationalism.
For almost two and a half millennia, reality was attributed by Western thinkers, and by
Western popular culture too, to elementary things, possessing fixed properties out of
which natural phenomena as experienced are composed. This finds expression in the
Published by Maney Publishing (c) IOM Communications Ltd

materialistic determinism of modern science, the analytic method developed by its

founders, the subject-predicate logic that dominated Western theories of reasoning
from Aristotle to the nineteenth century, and atomism in its many forms. The atomic
theory of matter, the cell theory in biology, the germ theory of disease, the gene theory
of inheritance, social atomism, all are instances of elementary unit thinking and modelling
of reality. The formulation in the last third of the nineteenth century of symbolic
logics of relations and field theories of electricity and magnetism first opened the way to
a reconceptualisation of reality by adding relationships to the list of the ultimately real.
The electromagnetic field is a relational structure that, though immaterial, is neverthe-
less a seat of forces. Durkheim conceptualised society as a network of relationships that
was also a seat of forces acting on individual members of society. Saussure did the same
for language, developing a theory of language as a closed system of relations. This system
is itself the source of all linguistic meanings and values in spite of language being a ‘social
fact’, hence intersubjective, and referring beyond itself, for example to the world. In
Einstein’s special and general theories of relativity, space and time are transformed from
Newtonian thinghood into relationships, and so are mass and energy. Concurrently, David
Hilbert championed interpreting mathematics as a network of logical relationships, and
Vilfredo Pareto modelled the economy relationally. In mid century, the anthropologist
Claude Levi-Strauss modelled cultures as relational structures, and in the 1960s and 70s
structuralism became a major explanatory tool in the humanities and social sciences.
Structuralism in the humanities and social sciences echoed one of the great discoveries
of nineteenth century chemistry, namely that structure itself could be a causal agent. In the
case of chemistry, this discovery was connected to the spatial arrangement of atoms within
a molecule. Molecules with exactly the same atomic constituents could have different
properties depending on the ‘stereometry’ of the molecule, the way those atoms were
organised. The importance of structure as a constituent of reality was rapidly assimilated
by physics, sociology, linguistics, economics and anthropology. To highlight one example,
X-ray crystallography, initially conceived as a way of proving that X-rays were electro-
magnetic waves, quickly became a tool for revealing molecular structure. Techniques for
crystallising proteins enabled Linus Pauling to discover the alpha helix structure of
proteins, and permitted Rosalind Franklin, inter alia, to generate the data necessary to
validate the Watson–Crick double helix model for the structure of DNA. Very quickly it
was discovered that the action of DNA was a function of the pattern of the four bases held

174 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2

 in place by the outer ‘shell’ of phosphates and sugars. It was this pattern alone that
distinguished one life form from another.
By the turn of the twenty-first century, study of the properties of structures had been
enhanced by the convergence of network theory, system level modelling of phenomena,
the study of self-organising, non-linear, far from equilibrium systems, and information
theory. This convergence has generated new relational theories of reality. There is, for
example, a growing recognition that genes and proteins act in and through extended gene,
protein and gene–protein networks, not only within individual cells but also among cell
networks. Cognitive neuroscience supports modelling the mind as activation patterns in
neuronal–glial networks. One approach to modelling a quantum theory of gravity involves
the idea that the universe itself is an information structure.25
Interpreting physical, biological and social realities in terms of networks is allied today
to increasingly sophisticated tools for studying and modelling networks. Networks are
relational structures with distinctive properties that are a function of the form of their
Published by Maney Publishing (c) IOM Communications Ltd

structure. This is highly suggestive of an approach to extending Dewey’s pragmatism by

identifying ‘objective’ values within experience modelled as a dynamical, evolving, far from
equilibrium relational structure. That experience is intrinsically normative is a corollary of
the view that action is motivated by desire. The problem has always been on what grounds
desires themselves can be judged. Historically, the dominant assumption in the West
has been that something fundamentally external to experience – God, Reality, the Good –
is necessary in order to provide such a ground. The pragmatist claim that a process inter-
pretation of experience can ground values in experience itself was provocative, but only
delivered a framework within which to critique means, not ends.
As engineering exemplifies a practice that successfully couples values and knowledge to
‘the world’, pursuing a philosophy of rational action by studying engineering practice
seems a particularly promising vehicle for exploring experience as itself a source of values.
Engineering surely is not uniquely qualified for this, but it has two claims on our attention.
First, engineering embodies a contingency based conception of rationality as a complex
process in which action and values are elementary features. This is an advance over any
conception of rationality that sets values and action aside as not within the scope of
rationality, knowledge and truth. Second, powerfully, there is a need today for rational
technology policies that would enable more effective technological action. Engineering is
now, and has for centuries been, ignored as a source of insight into the physical, social and
cultural problems associated with technological innovation. That all approaches favoured
by Western intellectuals continue to prove sterile seems to me to make a compelling case
for this to change.

NOTES
1. S. L. Goldman: ‘Philosophy, engineering and Western culture’, in Broad and Narrow Interpretations of
Philosophy of Technology, (ed. P. T. Durbin); 1990, Amsterdam, Kluwer.
2. Aristotle: Rhetoric, 1357a23–28.
3. V. Bush: Science: The Endless Frontier; 1980, Washington, DC, National Science Foundation Press.
4. C. Mitcham: Thinking Through Technology: The Path Between Engineering and Philosophy; 1999, Chicago, IL,
University of Chicago Press.
5. S. H. Cutcliffe: Ideas, Machines, and Values: An Introduction to Science, Technology, and Society Studies; 2000,
Lanham, MD, Rowman and Littlefield.
6. S. L. Goldman: ‘The social captivity of engineering’, in Critical Perspectives on Non-Academic Science and
Engineering, (ed. P. T. Durbin); 1991, Bethlehem, PA, Lehigh University Press.

INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2 175

 7. S. L. Goldman: ‘Images of science and technology in popular films’, Science, Technology and Human Values,
1989, 14, 275–301.
8. S. L. Goldman: ‘No innovation without representation: technological action in a democratic society’, in
New Worlds, New Ideas, New Issues, (ed. S. H. Cutcliffe et al.); 1992, Bethlehem, PA, Lehigh University Press.
9. S. L. Goldman: ‘The social captivity of engineering’ (see Note 6).
10. H. A. Simon: Models of Bounded Rationality, Vol. 3; 1997, Cambridge, MA, MIT Press.
11. C. Webster: The Great Instauration: Science, Medicine and Reform 1626–1660; 1976, New York, NY, Holmes
and Meier.
12. S. L. Goldman: ‘Philosophy, engineering and Western culture’ (see Note 1).
13. But see M. Untersteiner: The Sophists; 1954, New York, NY, Philosophical Library.
14. N. Struever: The Language of History in the Renaissance, 10; 1960, Princeton, NJ, Princeton University Press.
15. Plato: Cratylus, 386A–386E.
16. Aristotle: Nicomachaean Ethics, books 6 and 10.
17. Aristotle: Rhetoric, 1357a23–28, 1359b12.
18. Aristotle: Rhetoric, 1355a34.
19. R. Popkin: A History of Scepticism from Erasmus to Spinoza; 1986, Berkeley, CA, University of California Press;
C. Schmitt: Cicero Scepticus: A Study of the Influence of the Academica in the Renaissance; 1972, Hague, Martinus
Published by Maney Publishing (c) IOM Communications Ltd

Nijhoff.
20. N. Struever: The Language of History in the Renaissance, p. 37 (see Note 12).
21. D. Kelley: Foundations of Modern Historical Scholarship; 1970, New York, NY, Columbia University Press.
22. S. Y. Edgerton: The Heritage of Giotto’s Geometry: Art and Science on the Eve of the Scientific Revolution; 1991,
Ithaca, NY, Cornell University Press.
23. J. Dewey: The Quest for Certainty, 84; 1960, New York, NY, Capricorn Books.
24. J. Dewey: The Quest for Certainty, p. 3, also 8 (see Note 20).
25. L. Smolin: Three Roads to Quantum Gravity; 2003, New York, NY, Basic Books.

Steven L. Goldman (slg2@lehigh.edu) is Andrew W. Mellon Distinguished Professor in the Humanities at

Lehigh University, where he holds a joint appointment in the departments of philosophy and history. His
teaching and research centre on the history and philosophy of modern science and technology, and on the
social dynamics of contemporary technological innovation. He was for eleven years Director of Lehigh’s
Science, Technology and Society programme. Among a wide range of publications, he has authored or
coauthored influential technology policy reports for the US government, and four books on the implications
of innovation for strategic management. Since the mid 1980s he has been studying the role of engineering
in technological innovation.

176 INTERDISCIPLINARY SCIENCE REVIEWS, 2004, VOL. 29, NO. 2